Nature 424, 251 (17 July 2003) | doi:10.1038/424251b

Looking into the safety of AAV vectors

Mark A. Kay1 & Hiroyuki Nakai1

  1. Departments of Pediatrics and Genetics, Program in Human Gene Therapy, Stanford University, 300 Pasteur Drive, Stanford, California 94305-5208, USA


The News story "Harmful potential of viral vectors fuels doubts over gene therapy" (Nature 423, 573–574: 200310.1038/423573a) suggests that there is a reasonable probability that recombinant AAV vectors may cause or contribute to cancer in gene therapy subjects. As authors of the paper discussed in your story (H. Nakai et al. Nature Genet. 34, 297–302; 2003), we would like to emphasize that there is no evidence that AAV causes cancer in animals or humans, and that your concern is unfounded.

Our studies demonstrate that AAV vector DNA will preferentially integrate into active genes when delivered into the livers of mice. This has raised concerns because of recently published reports of leukaemia in two of nine patients treated with a recombinant retroviral vector for a lethal genetic disease, X-linked severe combined immunodeficiency disorder (SCID). The leukaemia was caused at least in part by the retroviral insertion and activation of an oncogene (insertional mutagenesis) in bone-marrow progenitor cells. Because retroviral vectors preferentially integrate into intragenic regions of the chromosome, your News story suggests that recombinant AAV vectors may pose similar risks in gene-therapy trials.

There are substantial differences between retroviral and AAV-mediated integration. First, unlike retroviral vectors, AAV-mediated vector integration is relatively uncommon. Second, retroviral vectors contain additional regulatory elements that are more likely than AAV vectors to activate a gene that they insert next to. Third, retroviral vectors contain the protein machinery needed to cause host chromosomal DNA breaks, whereas AAV does not. It is possible that AAV preferentially integrates into DNA regions that are already damaged within treated cells.

A symposium entitled "Safety considerations in the use of AAV vectors in gene transfer clinical trials", jointly sponsored by the NIH and the FDA, was held in March 2001 (see On the basis of data from hundreds of normal mice treated with the vector, the conclusion was reached that there was no evidence to suggest that the vector caused cancer.

In addition, the leukaemia found in patients treated with X-linked SCID gene therapy may be unique to this particular disease because of the unusual physiological events that occur after treatment. In X-linked SCID, the genetic reconstitution of a very few precursor cells results in the selective proliferation of immune cells genetically corrected with the vector. Any additional proliferation stimulus, such as the activation of an oncogene, may result in the further growth and expansion of these cells. This type of growth advantage is not a factor in most gene-therapy trials, and AAV has not been used in clinical trials to treat such disorders.

Although we support additional long-term safety studies, we believe the risk of cancer in current AAV trials is negligible, on the basis of infrequent integration efficiency and the quiescent nature of the target tissues.